EP1100373B1

EP1100373B1 - Intrabody navigation system for medical applications

Info

Publication number: EP1100373B1
Application number: EP19990929671
Authority: EP
Inventors: Pinhas Gilboa; Danny Blecher
Original assignee: SuperDimension Ltd
Current assignee: SuperDimension Ltd
Priority date: 1998-08-02
Filing date: 1999-07-07
Publication date: 2008-09-03
Anticipated expiration: 2019-07-07
Also published as: US6833814B2; JP2005128035A; DE69939471D1; US6593884B1; JP2003524443A; US20010047133A1; US20020042571A1; AU4644799A; US20030160721A1; EP2279692A3; US6947788B2; JP2005161077A; WO2000010456A1; EP1100373A4; JP2005161076A; US20020005719A1; JP2005185845A; US7555330B2; EP2100557A1; EP2279692A2

Abstract

This invention is a system, a method for tracking the position, and orientation of a probe such as a catheter whose transverse inner dimension may be at most about 2 millimeters. Three planar antennas (26, 28, 30) that at least partly overlap are used to transmit electromagnetic radiation simultaneously, with the radiation transmitted by each antenna (26, 28, 30) having its own spectrum. In the case of single-frequency spectra, the antennas (26, 28, 30) are provided with mechanisms for decoupling them from each other. A receiver (10) inside the probe includes sensors (16a, 16b, 18a, 18b, 20a, 20b) of the three components of the transmitted field, with sensors (16a, 16b, 18a, 18b, 20a, 20b) for at least two of the three components being pairs of sensors (16a, 16b, 18a, 18b, 20a, 20b), such as coils, disposed symmetrically with respect to a common reference point.

Description

FIELD AND BACKGROUND OF THE INVENTION

The present invention relates to electromagnetic tracking devices and, more particularly, to a system for tracking a medical probe such as a catheter as the probe is moved through the body of a patient.
It is known to track the position and orientation of a moving object with respect to a fixed frame of reference, by equipping the moving object with a transmitter that transmits electromagnetic radiation, placing a receiver in a known and fixed position in the fixed frame of reference, and inferring the continuously changing position and orientation of the object from signals transmitted by the transmitter and received by the receiver. Equivalently, by the principle of reciprocity, the moving object is equipped with a receiver, and a transmitter, is placed in a known and fixed position in the fixed frame of reference. Typically, the transmitter includes three orthogonal magnetic dipole transmitting antennas; the receiver includes three orthogonal magnetic dipole receiving sensors; and the object is close enough to the stationary apparatus (transmitter or receiver), and the frequencies of the signals are sufficiently low, that the signals are near field signals. Also typically, the system used is a closed loop system: the receiver is hardwired to, and explicitly synchronized with, the transmitter. Representative prior art patents in this field include US 4,287,809 and US 4,394,831, to Egli et al. ; US 4,737,794, to Jones ; US 4,742,356, to Kuipers ; US 4,849,692, to Blood ; and US 5,347,289, to Elhardt. Several of the prior art patents, notably Jones, present non-iterative algorithms for computing the position and orientation of magnetic dipole transmitters with respect to magnetic dipole receivers.
An important variant of such systems is described in US 5,600,330, to Blood. In Blood's system, the transmitter is fixed in the fixed reference frame, and the receiver is attached to the moving object. Blood's transmitting antennas are spatially extended, and so cannot be treated as point sources. Blood also presents an algorithm which allows the orientation, but not the position, of the receiver relative to the transmitter to be calculated non-iteratively.
Systems similar to Blood's are useful for tracking a probe, such as a catheter or an endoscope, as that probe is moved through the body of a medical patient. It is particularly important in this application that the receiver be inside the probe and that the transmitter be external to the patient, because transmitting antennas of sufficient power would not fit inside the confined volume of the probe. A representative prior art system of this type is described in PCT Publication WO 96/05768 , entitled "Medical Diagnosis, Treatment and Imaging System". Medical applications of such systems include cismyocardial revascularization, balloon catheterization, stent emplacement, electrical mapping of the heart and the insertion of nerve stimulation electrodes into the brain.
Perhaps the most important application of this tracking is to intrabody navigation, as described by Acker in US Patent No. 5,729,129 , with reference to PCT Publication No. WO 95/09562 . A three-dimensional image, such as a CT or MRI image, of the patient is acquired. This image includes fiducial markers at predetermined fiducial points on the surface of the patient. Auxiliary receivers similar to the receiver of the probe are placed at the fiducial points. The signals received by the auxiliary receivers are used to register the image with respect to the transmitter frame of reference, so that an icon that represents the probe can be displayed, superposed on a slice of the image, with the correct position and orientation , with respect to the image. In this way, a physician can see the position and orientation of the probe with respect to the patient's organs.
WO 96/05768 illustrates Another constraint imposed on such systems by the small interior dimensions of the probe. In most prior art systems, for example, the system of Egli et al., the receiver sensors are three concentric orthogonal coils wound on a ferrite core. The coils are "concentric" in the sense that their centers coincide. Such a receiver of sufficient sensitivity would not fit inside a medical probe. Therefore, the sensor coils of WO 96/05768 are collinear: the three orthogonal coils are positioned one behind the other, with their centers on the axis of the probe, as illustrated in Figure 3 of WO 96/05768 . This reduces the accuracy of the position and orientation measurements, because instead of sensing three independent magnetic field components at the same point in space, this receiver senses three independent magnetic field components at three different, albeit closely spaced, points in space.
A further, consequent concession of the system of WO 96/05768 to the small interior dimensions of a catheter is the use of coils wound on air cores, rather than the conventional ferrite cores. The high mutual coupling of collinear coils wound on ferrite cores and measuring three independent field components at three different points in space would distort those measurements sufficiently to make those measurements fatally nonrepresentative of measurements at a single point.
Another drawback of the system of WO 96/05768 relates to the geometry of the transmitter antennas. These are three nonoverlapping flat coplanar coils, preferably arranged in a triangle: Because the strength of the field transmitted by one of these coils falls as the reciprocal cube of the distance from the coil, the receiver usually senses fields of very disparate strength, which further degrades the accuracy of the position and orientation measurements. Acker addresses this problem by automatically boosting the power supplied to transmitting coils far from the receiver. In US Patent No. 5,752,513 , Acker et al. address this problem by overlapping the coplanar transmitting coils.
Acker et al. transit time-multiplexed DC signals. This time multiplexing slows drown the measurement. Frequency multiplexing, as taught in WO 96/05768 , overcomes this problem, but introduces a new problem insofar as the transmitting coils are coupled by mutual inductance at non-zero transmission frequency, so that the transmitted field geometry is not the simple geometry associated with a single coil, but the more complex geometry associated with several coupled coils. This complicates and slows down the calculation of the position and orientation of the receiver relative to the transmitter coils. PCT Publication WO 97/36143 , entitled "Mutual Induction Correction", addresses this problem by generating, at each transmitter coil, counter-fields that cancel the fields generated by the other transmitter coils.
A further source of slowness in calculating the position and orientation of the receiver is the iterative nature of the calculation required for a spatially extended transmitter. As noted above, Blood calculates the position of the receiver iteratively. Even in the DC case, Acker et al. calculate both the position and the orientation of the receiver iteratively.
There is thus a widely recognized need for, and it would be highly advantageous to have, a faster and more accurate method for tracking a medical probe inside the body of a patient.
WO 97/29684 discloses a catheter with helical coils, for sensing electromagnetic field components, that are arranged symmetrically with respect to a common reference point in order to accommodate a lumen.

SUMMARY OF THE INVENTION

According to the present invention there is provided a system for tracking as defined in claim 1,

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is herein described, by way of example only, with reference to the accompanying drawings, wherein:

FIG 1 is a schematic diagram of a system useful for understanding the present invention;
FIG. 2A is a partly cut away perspective view of a probe and a receiver;
FIG. 2B is a circuit diagram of the receiver of FIG. 2A;
FIG. 2C illustrates features of the receiver of FIG. 2A that suppress unwanted electromagnetic coupling;
FIG. 3 is an axial sectional view of a probe and a receiver;
FIG. 4A shows two coils of opposite helicities;
FIG. 4 shows two coils of identical helicities;
FIG. 5 shows a second example of an embodiment of a receiver;
FIG. 6 is a plan view of three loop antennas and two phantom loop antennas;
FIGs. 7A, 7B and 7C show alternative configurations of paired adjacent loop antennas;
FIG. 8 is a schematic block diagram of driving circuitry
FIG. 9 shows, a C-mount fluoroscope modified for real-time intrabody navigation
FIG. 10 shows a coil of the receiver of FIG. 5;
FIG. 11 shows a CT scanner modified for imaging in support of subsequent intracranial navigation;
FIG. 12A is a partly cut-away perspective view of a cardiac catheter of the present invention in the retracted position thereof;
FIG. 12B is a perspective view of the catheter of FIG. 12A in the extended position thereof;
FIG. 12C is an end-on view of the catheter of FIG. 12a in the retracted position thereof;
FIG. 13A is a partly cut-away side view of an example of the cardiac catheter in the retracted and inflated position thereof;
FIG. 13B is an end-on view of the catheter of FIG. 13A in the retracted and inflated position thereof;
FIG. 14 is a partial perspective view of the C-mount fluoroscope of FIG. 9, including a magnetically permeable compensator;
FIG. 15 is a partial exploded perspective view of a preferred embodiment of the probe and receiver of FIG. 2A;
FIG. 16 illustrates a scheme for retrofitting an apparatus such as the receiver of FIG. 2A to a catheter.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention is of a system for tracking the position and orientation of an object relative to a fixed frame of reference. Specifically, the present invention can be used to track the motion of a medical probe such as a catheter or an endoscope within the body of a patient.
The principles and operation of remote tracking according to the present invention may be better understood with reference to the drawings and the accompanying description.
Referring now to the drawings, Figure 1 illustrates, in general terms, a system for understanding the present invention. Within a probe 10 is rigidly mounted a receiver 14. Receiver 14 includes three field component sensors 16, 18, and 20, each for sensing a different component of an electromagnetic field. Sensor 16 includes two sensor elements 16a and 16b. Sensor 18 includes two sensor elements 18a and 18b. Sensor 20 includes two sensor elements 20a and 20b. Typically, the sensor elements are coils, and the sensed components are independent magnetic field components. Sensor elements 16a and 16b arc on opposite sides of, and equidistant from, a common reference point 22. Similarly, sensor elements 18a and 18b are on opposite sides of, and equidistant from, point 22, and sensor elements 20a and 20b also are on opposite sides of, and equidistant from, point 22. In the illustrated example, sensors 16, 18 and 20 are disposed collinearly along a longitudinal axis 12 of probe 10, but other configurations are possible, as discussed below.
The system of Figure 1 also includes a transmitter 24 of electromagnetic radiation. Transmitter 24 includes three substantially coplanar rectangular loop antennas 26, 28 and 30 connected to driving circuitry 32. Loop antennas 26 and 28 are adjacent and are partly overlapped by loop antenna 30. Driving circuitry 32 includes appropriate signal generators and amplifiers for driving each of loop antennas 26, 28 and 30 at a different frequency. The electromagnetic waves generated by transmitter 24 are received by receiver 14. The signals from receiver 14 that correspond to these electromagnetic waves are sent to reception circuitry 34 that includes appropriate amplifiers and A/D converters. Reception circuitry 34 and driving circuitry 32 are controlled by a controller/processor 36 that typically is an appropriately programmed personal computer. Controller/processor 36 directs the generation of transmitted signals by driving circuitry 32 and the reception of received signals by reception circuitry 34. Controller/processor 36 also implements the algorithm described below to infer the position and orientation of probe 10. Note that the system of Figure 1 is a closed-loop system: the reception of signals from receiver 14 is synchronized with the transmission of electromagnetic waves by transmitter 24
Figure 2 shows an embodiment of receiver 14 according to the invention. Figure 2A is a perspective, partly cut away view of probe 10 with receiver 14 mounted in the housing 11 thereof. Figure 2B is a circuit diagram of receiver 14. In this embodiment, sensor elements 16a, 16b, 18a and 18b are coils of conducting wire wound on ferrite cores 70. Coils 16a and 16b are mutually parallel. Coils 18a and 18b are mutually parallel and are perpendicular to coils 16a and 16b. Coils 16a, 16b, 18a and 18b all are perpendicular to axis 12. Instead of sensor 20 with two sensor elements 20a and 20b, the embodiment of Figure 2 has a single coil 20' of conducting wire wound on a ferrite core 70. Coil 20' is parallel to axis 12 and therefore is perpendicular to coils 16a, 16b, 18a and 18b. Coil 20' is centered on reference point 22. Sensors 16, 18 and 20' are connected to reception circuitry 34 by twisted wire pairs 38. As shown in the circuit diagram of Figure 2B. coils 16a and 16b are connected in series, and coils 18a and 18b are connected in series.
Because sensors 16, 18 and 20' of Figure 2 all measure field components at the same reference point 22, coils 16a, 16b, 18a, 18b and 20' can be wound on ferrite cores 70 instead of the air cores of WO 96/05768 without causing undue distortion of the received signals, despite the small transverse interior diameter 72, typically less than two millimeters, of probe 10 when probe 10 is a catheter.
Wire pairs 38 are twisted in order to suppress electromagnetic coupling between wire pairs 38 and the environment, and in particular to suppress electromagnetic coupling between wire pairs 38 and transmitter 24. Figure 2C is a circuit diagram that shows further features of the present invention that suppress this electromagnetic coupling. Figure 2C is drawn with particular reference to sensor 16, but the same features apply, mutatis mutandis, to sensor 18.
Coils 16a and 16b are connected in series by inner leads 116a and 116b thereof. Outer lead 216a of coil 16a is connected, by wire 38a of twisted wire pair 38, to a positive input 126a of a differential amplifier 128 of reception circuitry 34. Outer lead 216b of coil 16b is connected, by wire 38b of twisted wire pair 38, to a negative input 126b of differential amplifier 128. Inner leads 116a and 116b also are connected to ground 124 by a wire 122. For illustrational clarity, wire 38a is drawn as a solid line, wire 38b is drawn as a dotted line and wire 122 is drawn as a dashed line.
Figure 15 is a partial exploded perspective view of a preferred embodiment of probe 10 and receiver 14. Housing 11 is substantially cylindrical, with two recesses 511 and 513 incised therein. The boundary of each recess 511 or 513 includes a pair of diametrically opposed apertures: apertures 510 and 512 in the boundary of recess 511 and apertures 514 and 516 in the boundary of recess 513. Arrows 530 and 532 show two of the three components of a cylindrical coordinate system for describing position within and along housing 11. Arrow 530 points in the longitudinal direction. Arrow 532 points in the azimuthal direction. Aperture pair 510, 512 is displaced both longitudinally and azimuthally from aperture pair 514, 516.
Coil 16a is a coil of electrically conducting wire that is wound about a core 70a. Core 70a is mounted in apertures 514 and 516: end 518 of core 70a, that extends beyond coil 16a, is mounted in aperture 514 and is secured rigidly in place by a suitable glue, and end 520 of core 70a, that extends beyond coil 16a in the opposite direction, is mounted in aperture 516 and is secured rigidly in place by a suitable glue. Similarly, coil 18a is a coil of electrically conducting wire that is wound about a core 70b. Core 70b is mounted in apertures 510 and 512: end 522 of core 70b, that extends beyond coil 18a, is mounted in aperture 510 and is secured rigidly in place by a suitable glue, and end 524 of core 70b, that extends beyond coil 18a in the opposite direction, is mounted in aperture 512 and is secured rigidly in place by a suitable glue.
Figure 15 also shows the preferred azimuthal separation of aperture pair 514, 516 from aperture pair 510, 512. Aperture pair 514, 516 is perpendicular to aperture pair 510, 512, in the sense that aperture pair 514, 516 is displaced 90°, in the direction of arrow 532, from aperture pair 510, 512. This makes core 70a perpendicular to core 70b, and hence makes coils 16a and 18a mutually perpendicular.
In the case of probe 10 being a catheter for invasively probing or treating a body cavity such as a chamber of the heart, it is preferable that housing 11 be made of a nonmagnetic metal such as nitinol, titanium, iconel, phynox or stainless steel. Housing 11 thus is sufficiently flexible to bend under the lateral forces of the walls of blood vessels through which probe 10 is inserted towards the body cavity, and sufficiently resilient to return to its unstressed shape, with coils 16a and 18a mutually perpendicular, when the portion of probe 10 that includes receiver 14 reaches the interior of the body cavity. Surprisingly, it has been found that the use of a conductive metal as the material of housing 11 does not distort the electromagnetic field sensed by receiver 14 despite the current eddies induced in housing 11 by the electromagnetic waves generated by transmitter 24. Apertures 510, 512, 514 and 516 are most conveniently formed by laser cutting. The accuracy of the mutual perpendicularity of coils 16a and 18a obtained in this manner has been found to be superior to the accuracy obtained by forming housing 11 as a solid cylindrical block and drilling mutually perpendicular recesses in the block to receive coils 16a and 18a.
Coils 16b and 18b are mounted similarly in similar pairs of diametrically opposed, azimuthally and longitudinally displaced apertures. This ensures that coils 16a and 16b are mutually parallel, that coils 18a and 18b are mutually parallel, and that coils 16b and 18b are mutually perpendicular.
In an alternative structure (not shown) of housing 11, housing 11 is formed as an open, spring-like frame that includes apertures 510, 512, 514 and 516 in the form of small rings that are sized to accept the ends 518, 520, 522 and 524 of cores 70a and 70b. The spring-like nature of this embodiment of housing 11 allows coils 16a and 18a to be mounted therein simply by forcing ends 518, 520, 522 and 524 into their respective apertures, and also allows housing 11 to flex during insertion towards a body cavity of a patient and to return to its unstressed shape upon arrival inside the body cavity.
Figure 3 is an axial sectional view of receiver 14 mounted in a variant of probe 10 that has two sections 10a and 10b connected by a flexible connector 40. As in Figure 2, sensors 16 and 18 include sensor elements 16a, 16b, 18a and 18b that are coils of conducting wire wound on air cores and that are perpendicular to axis 12. Sensor elements 16a, and 16b are mutually parallel, sensor elements 18a and 18b are mutually parallel, and sensor elements 16a and 16b are perpendicular to sensor elements 18a and 18b. Sensor 20 includes two sensor elements: coils 20a and 20b of conducting wire wound on air cores. Coils 20a and 20b are equidistant from reference point 22 and are parallel to axis 12. Like coils 16a and 16b and like coils 18a and 18b, coils 20a and 20b are connected in series. Flexible connector 40 allows this variant of probe 10 to bend as this variant of probe 10 is moved within a medical patient. Sensor element pairs 16, 18 and 20 are disposed symmetrically with respect to reference point 22 in the sense that when probe 10 of Figure 3 is straight, as drawn, sensor elements 16a and 16b are on opposite sides of, and equidistant from, reference point 22; and likewise sensor elements 18a and 18b are on opposite sides of, and are equidistant from, reference point 22; and sensor elements 20a and 20b are on opposite sides of, and are equidistant from, reference point 22. Note that when probe 10 of Figure 3 is straight, sensor elements 16a, 16b, 18a, 18b, 20a and 20b all are collinear, along axis 12 that intersects point 22, and so are disposed symmetrically with respect to point 22.
For coil pairs such as pairs 16a and 16b to produce signals representative of a magnetic field component at point 22 when the coil pairs are connected as shown in Figure 2A, the two coils must have opposite helicity; as illustrated in Figure 4A, so that, in a spatially uniform time varying magnetic field, the signals induced in the two coil pairs 16a and 16b reinforce each other instead of canceling each other. Coil pairs 16a and 16b that have identical helicities, as illustrated in Figure 4B, may be used to measure a magnetic field component gradient at point 22. Alternatively, coil pairs of identical helicities may be used to measure magnetic field components if the top of one coil is connected to the bottom of the other coil.
Figure 5 illustrates a second class of an example of receiver 14. In Figure 5, a conceptual cylindrical surface is denoted by dashed lines 42 and dashed circles 44. The embodiment of receiver 14 illustrated in Figure 5 includes three sensors 16, 18 and 20, each with two sensor elements 16c and 16d, 18c and 18d, and 20c and 20d, respectively. Each sensor element is a flat rectangular coil, of many turns of conducting wire, that is bent into an arcuate shape to conform to the shape of the cylindrical surface. Sensor elements 16c, 18c and 20c are interleaved around circle 44a. Sensor elements 16d, 18d and 20d are interleaved around circle 44b. Sensor elements 16c and 16d are disposed symmetrically with respect to reference point 22, meaning that sensor elements 16c and 16d are on opposite sides of reference point 22, are equidistant from reference point 22, and are oriented so that an appropriate 180° rotation about point 22 maps sensor 16c into sensor 16d. Similarly, sensor elements 18c and 18d are disposed symmetrically with respect to reference point 22, and sensor elements 20c and 20d are disposed symmetrically with respect to reference point 22. Sensor elements 16c and 16d are connected in series, in a manner similar to sensor elements 16a and 16b, to respond to one component of the magnetic field. Sensor elements 18c and 18d are connected similarly in series to respond to a second component of the magnetic field that is independent of the first component, and sensor elements 20c and 20d are connected similarly in series to respond to a third component of the magnetic field that is independent of the first two components. Most preferably, sensor elements 16c, 16d, 18c, 18d, 20c and 20d are sized and separated so that these three magnetic field components are orthogonal. In practice, the cylindrical surface whereabout sensor elements 16c, 16d, 18c, 18d, 20c and 20d are disposed could be the inner surface of probe 10 or the outer surface of a cylindrical sleeve adapted to fit inside probe 10. In the case of this embodiment of receiver 14 formed on the outer surface of a cylindrical sleeve, sensor elements 16c, 16d, 18c, 18d, 20c and 20d may be fabricated by any one of several standard methods, including photolithography and laser trimming. Figure 10 illustrates the preferred geometry of sensor elements 16c, 16d, 18c, 18d, 20c and 20d: a flat rectangular spiral 17 of an electrical conductor 19. Only four turns are shown in spiral 17, for illustrational simplicity. Preferably, however, there are several hundred turns in spiral 17. For example, a spiral 17, intended for a cylindrical surface of a diameter of 1.6 millimeters, in which conductor 19 has a width of 0.25 microns, and in which the windings are separated by gaps of 0.25 microns, has 167 turns.
Figures 12A, 12B and 12C illustrate the distal end of a cardiac catheter 300. Figure 12A is a partly cut-away perspective view of catheter300 in the retracted position thereof. Figure 12B is a perspective view of catheter 300 in the extended position thereof. Figure 12C is an end-on view of catheter 300 in the retracted position thereof. Catheter 300 includes a flexible cylindrical inner sleeve 302 slidably mounted in a flexible cylindrical outer sleeve 304. Connecting distal end 306 of inner sleeve 302 to distal end 308 of outer sleeve 304 are four flexible rectangular strips 310. When inner sleeve 302 is in the extended position thereof relative to outer sleeve 304, strips 310 are flush against inner sleeve 302, as shown in Figure 12B. When inner sleeve 302 is in the retracted position thereof relative to outer sleeve 304, strips 310 bow outward in circular arcs, as shown in Figure 12A.
Catheter 300 includes a set of three orthogonal electromagnetic field component sensors 316, 318 and 320, in the manner of receiver 14 of Figure 1. First sensor 316 includes coils 316a and 316b mounted on opposite lateral edges 312a and 314a of strip 310a and on opposite lateral edges 312c and 314c of strip 310c. Coil 316a is mounted on lateral edges 312a and 312c. Coil 316b is mounted on lateral edges 314a and 314b. Second sensor 318 includes coils 318a and 318b mounted on opposite lateral edges 312b and 314b of strip 310b and on opposite lateral edges 312d and 314d of strip 310d. Coil 318a is mounted on lateral edges 312b and 312d. Coil 318b is mounted on lateral edges 314b and 314d. Third sensor 320 includes coils 320a and 320b. Inner sleeve 302 is cut away in Figure 12A to show coils 320a and 320b. For illustrational clarity, the wires of coils 316a and 318a are shown in Figures 12A and 12B as dashed lines, and only two turns are shown for each coil, although in practice at least nine turns of 45-micron-diameter copper wire are used. Note that the wires of coil 316a run through inner sleeve 302, from lateral edge 312a to lateral edge 312c, and do not terminate at the intersection of lateral edges 312a and 312c with inner sleeve 302. Similarly, the wires of coil 318a do not terminate at the intersection of lateral edges 312b and 312d with inner sleeve 302, but instead run from lateral edge 312b to lateral edge 312d. Also for illustrational clarity, lateral edges 312 are shown much wider than they really are in catheter 300. Coils 320a and 320b are wound around a permeable core (not shown).
In a catheter 300, the length of inner sleeve 302 exceeds the length of outer sleeve 304 by 15.7mm in the extended position. Also in a typical embodiment of catheter 300, each of coils 320a and 320b is about 1.1mm long and about 1.1mm in diameter and includes about 400 turns of 10 micron diameter copper wire.
Coils 320a and 320b are parallel and equidistant from a central point 322. When catheter 300 is opened to the retracted position thereof, as shown in Figures 12A and 12C, the circular arcs formed by strips 310 are concentric with point 322. This makes coils 316a, 316b, 318a and 318b circular and concentric with point 322, with coils 316a and 316b being mutually parallel, and with coils 318a and 318b being mutually parallel, so that point 322 then becomes the reference point for electromagnetic field measurements.
In the extended position thereof, catheter 300 is thin enough, preferably less than about 2mm in diameter, to be inserted via the blood vessels of a patient into the patient's heart: Once the distal end of catheter 300 is inside the desired chamber of the patient's heart, inner sleeve 302 is withdrawn relative to outer sleeve 304 to put catheter 300 in the retracted position thereof. Sensors 316, 318 and 320 are used in conjunction with transmitter 24 in the manner described below to determine the location and orientation of the distal end of catheter 300 within the patient's heart.
Mounted on outward faces 324 of strips 310 are four electrodes 326. Mounted on distal end 306 of inner sleeve 302 is an electrode 328. Electrodes 326 and 328 may be used for electrophysiologic mapping of the patient's heart. Alternatively, high RF power levels may be applied to selected heart tissue via electrode 328 to ablate that tissue in the treatment of conditions such as ventricular tachycardia.
Figures 13A and 13B illustrate the distal end of an alternative embodiment 400 of the cardiac catheter of the present invention. Figure 13A is a partly cut-away side view of catheter 400 in the retracted position thereof. Figure 13B is an end-on view of catheter 400 in the retracted position thereof. Like catheter 300, catheter 400 includes a flexible cylindrical inner sleeve 402 slidably mounted in a flexible cylindrical outer sleeve 404. Connecting distal end 406 of inner sleeve 402 to distal end 408 of outer sleeve 404 is a single flexible member: an inflatable latex balloon 410. When inner sleeve 402 is in the extended position thereof relative to outer sleeve 404, balloon 410 is flush against inner sleeve 402. After the illustrated distal end of catheter 400 has been introduced to the targeted chamber of a patient's heart, inner sleeve 402 is withdrawn to the retracted position thereof, and balloon 410 is inflated to assume a spherical shape.
Like catheter 300, catheter 400 includes a set of three orthogonal electromagnetic field component sensors 416, 418 and 420, in the manner of receiver 14 of Figure 1. First sensor 416 includes parallel coils 416a and 416b mounted as shown on outer surface 412 of balloon 410. Second sensor 418 includes parallel coils 418a and 418b mounted orthogonally to coils 416a and 416b on outer surface 412, as shown. Third sensor 420 includes coils 420a and 420b. Balloon 410 and inner sleeve 402 are cut away in Figure 13A to show coils 420a and 420b. Coils 420a and 420b are parallel and equidistant from a central point 422. When catheter 400 is opened to the retracted position thereof and balloon 410 is inflated to a spherical shape, outer surface 412 is a sphere concentric with point 422. This makes coils 416a, 416b, 418a and 418b circular and concentric with point 422, so that point 422 then becomes the reference point for electromagnetic field measurements.
Also as in the case of catheter 300, catheter 400 includes four electrodes 426, similar to electrodes 326, mounted on outer surface 412, and an electrode 428, similar to electrode 328, mounted on distal end 406 of inner sleeve 402.
Figure 6 is a plan view of loop antennas 26, 28 and 30. Loop antenna 26 is a rectangle with legs 26a, 26b, 26c and 26d. Loop antenna 28 is a rectangle of the same shape and size as loop antenna 26, and with legs 28a, 28b, 28c and 28d. Legs 26b and 28d are adjacent. Loop antenna 30 also is rectangular, with legs 30a, 30b, 30c and 30d. Leg 30a overlies legs 26a and 28a; leg 30b overlies the upper half of leg 28b; and leg 30d overlies the upper half of leg 26d, so that loop antenna 30 overlaps half of loop antenna 26 and half of loop antenna 28. Also shown in phantom in Figure 6 is a fourth rectangular loop antenna 46 and a fifth rectangular loop antenna 48 that are not part of transmitter 24 but are referred to in the explanation below. Loop antenna 46 is of the same shape and size as loop antenna 30, and overlaps the halves of loop antennas 26 and 28 that are not overlapped by loop antenna 30. Loop antenna 48 matches the outer perimeter defined by loop antennas 26 and 28.
To understand the preferred mode of the operation of the system of the present invention, it is helpful to consider first a less preferred mode, based on time domain multiplexing, of operating a similar system that includes all five loop antennas of Figure 6. In this less preferred mode, loop antenna 48 is energized using a sinusoidal current of angular frequency ω₁. Then, loop antennas 26 and 28 are energized by oppositely directed sinusoidal currents of angular frequency ω₁. Finally, loop antennas 30 and 46 are energized by oppositely directed sinusoidal currents of angular frequency ω₁. The idea of this energization sequence is to produce, first, a field above the transmitter that is spatially symmetric in both the horizontal and the vertical direction as seen in Figure 6, then a field above the transmitter that is antisymmetric in the horizontal direction and symmetric in the vertical direction, and finally a field that is symmetric in the horizontal direction and antisymmetric in the vertical direction. These three fields are linearly independent, and all three fields have significant amplitude all the way across the transmitter. The signals output by the three sensors of receiver 14 in response to the electromagnetic waves so generated are sampled at times t_m by reception circuitry 34. The sampled signals are: ${s^{0}}_{im} = {c^{0}}_{i, 1} {cosω}_{1} t_{m} + {c^{0}}_{i, 2} {sinω}_{1} t_{m}$
from loop antenna 48 ${s^{h}}_{im} = {c^{h}}_{i, 1} {cosω}_{1} t_{m} + {c^{h}}_{i, 2} {sinω}_{1} t_{m}$
from loop antennas 26 and 28 ${s^{v}}_{im} = {c^{v}}_{i, 1} {cosω}_{1} t_{m} + {c^{v}}_{i, 2} {sinω}_{1} t_{m}$
from loop antennas 30 and 46
where i indexes the sensor that receives the corresponding signal. Coefficients c ⁰ _i,l, c^h _i,l and c^v _i,l are the in-phase amplitudes of the received signals. Coefficients c ⁰ _i,2, c^h _i,2 and c^v _i,2 are the quadrature amplitudes of the received signals. Because ω₁ is sufficiently low that receiver 14 is in the near fields generated by the loop antennas, in principle the quadrature amplitudes should be identically zero. Because of inevitable phase distortions, for example in reception circuitry 34, the quadrature amplitudes generally are not zero.
Note that amplitudes c ⁰ _i,j, c^h _i,j and c^v _i,j (j=1,2) could be obtained by using only loop antennas 26, 28 and 30. The sampled signals obtained by energizing loop antennas 26, 28 and 30 separately with identical sinusoidal currents of angular frequency ω₁ are: ${s^{1}}_{im} = {c^{1}}_{i} {cosω}_{1} t_{m} + {c^{2}}_{i} {sinω}_{1} t_{m}$
from loop antenna 26 ${s^{2}}_{im} = {c^{3}}_{i} {cosω}_{1} t_{m} + {c^{4}}_{i} {sinω}_{1} t_{m}$
from loop antenna 28 ${s^{3}}_{im} = {c^{5}}_{i} {cosω}_{1} t_{m} + {c^{6}}_{i} {sinω}_{1} t_{m}$
from loop antenna 30
the coefficients c ¹ _i , c³ _i and c ⁵ _i being in-phase amplitudes and the coefficients c ² _i , c ⁴ _i and c ⁶ _i being quadrature amplitudes. Because the field radiated by loop antennas 26 and 28 when identical currents J flow therein is the same as the field generated by loop antenna 48 when current J flows therein, ${c^{0}}_{i, 1} = {c^{1}}_{i} + {c^{3}}_{i}$
${c^{0}}_{i, 2} = {c^{2}}_{i} + {c^{4}}_{i}$

By definition, ${c^{h}}_{i, 1} = {c^{1}}_{i} - {c^{3}}_{i}$
${c^{h}}_{i, 2} = {c^{2}}_{i} - {c^{4}}_{i}$

Finally, the fact that the field radiated by loop antenna 48 could also be emulated by identical currents flowing through loops 30 and 46 gives ${c^{v}}_{i, 1} = 2 {c^{5}}_{i} - {c^{1}}_{i} + {c^{3}}_{i}$
${c^{v}}_{i, 2} = 2 {c^{6}}_{i} - {c^{2}}_{i} - {c^{4}}_{i}$
In the preferred mode of the operation of the system of the present invention, loop antennas 26, 28 and 30 are energized simultaneously with sinusoidal currents of angular frequencies ω₁, ω₂ and ω₃, respectively. The sampled signals now are S _im = c _i1cosω₁ t_m + c _i2sinω₁ t_m + c _i3cosω₂ t_m + c_i4 sinω₂ t_m + c_i5cosω₃ t_m + c_i6 sinω₃ t_m (7) Note that now, amplitudes c _i1 and c _i2 refer to frequency ω_1, amplitudes c _i3 and c _i4 refer to frequency ω₂, and amplitudes c_i5 and c_i6 refer to frequency ω₃. The sampled signals are organized in a matrix s of three rows, one row for each sensor of receiver 14, and as many columns as there are times t_m, one column per time. Amplitudes c_ij are organized in a matrix c of three rows and six columns. The matrices s and c are related by a matrix A of six rows and as many columns as there are in matrix s: $s = cA$

Almost always, there are many more than six columns in matrix s, making equation (8) highly overdetermined. Because the transmission frequencies and the reception times are known, matrix A is known. Equation (8) is solved by right-multiplying both sides by a right inverse of matrix A: a matrix, denoted as A^-1 , such that AA^-1 = I, where I is the 6x6 identity matrix. Right inverse matrix A^-1 is not unique. A particular right inverse matrix A^-1 may be selected by criteria that are well known in the art. For example, A^-1 may be the right inverse of A of smallest L² norm. Alternatively, matrix c is determined as the generalized inverse of equation (8): $c = {sA}^{T} ({AA}^{T})^{- 1}$

where the superscript "T" means "transpose". The generalized inverse has the advantage of being an implicit least squares solution of equation (8).
In the special case of evenly sampled times t_m , solving equation (8) is mathematically equivalent to the cross-correlation of WO 96/05768 . Equation (8) allows the sampling of the signals from receiver 14 at irregular times. Furthermore, there is no particular advantage to using frequencies ω₁, ω₂ and ω₃ that are integral multiples of a base frequency. Using closely spaced frequencies has the advantage of allowing the use of narrow-band filters in reception circuitry 34, at the expense of the duration of the measurement having to be at least about 2π/Δω, where Δω is the smallest frequency spacing, except in the special case of two signals of the same frequency and different phases.
Because receiver 14 is in the near field of transmitter 24, coefficients c_ij of equation (7) are the same as coefficients c^j _i . It follows that equations (1)-(6) still hold, and either of two 3x3 matrices M can be formed from the elements of matrix c for further processing according to the description in co-pending Israel Patent Application 122578 , an in-phase matrix $M = (\begin{matrix} c_{1, 1}^{0} & c_{1, 1}^{h} & c_{1, 1}^{v} \\ c_{2, 1}^{0} & c_{2, 1}^{h} & c_{2, 1}^{v} \\ c_{3, 1}^{0} & c_{3, 1}^{h} & c_{3, 1}^{v} \end{matrix})$

or a quadrature matrix $M = (\begin{matrix} c_{1, 2}^{0} & c_{1, 2}^{h} & c_{1, 2}^{v} \\ c_{2, 2}^{0} & c_{2, 2}^{h} & c_{2, 2}^{v} \\ c_{3, 2}^{0} & c_{3, 2}^{h} & c_{3, 2}^{v} \end{matrix})$

Note that because the system is a closed-loop system, there is no sign ambiguity in M, unlike the corresponding matrix of co-pending Israel Patent Application 122578 .
Let T be the orthonormal matrix that defines the rotation of probe 10 relative to the reference frame of transmitter 24. Write M in the following form: $M = {ET}_{0} T$

where T₀ is an orthogonal matrix and E is in general a nonorthogonal matrix. In general, T ₀ and E are functions of the position of probe 10 relative to the reference frame of transmitter 24. Let $W^{2} = {MM}^{T} = {ET}_{0} {TT}^{T} {T_{0}}^{T} E^{T} = {EE}^{T}$

W ² is real and symmetric, and so can be written as W² = Pd²P^T = (PdP^T)², where d² is a diagonal matrix whose diagonal elements are the (real and positive) eigenvalues of W ² and where P is a matrix whose columns are the corresponding eigenvectors of W² Then W = PdP^T = E also is symmetric. Substituting in equation (12) gives: $M = {PdP}^{T} T_{0} T$

so that $T = {T_{0}}^{T} {Pd}^{- 1} P^{T} M$

If T ₀ is known, then T , and hence the orientation of probe 10 with respect to the reference frame of transmitter 24, can be computed using equation (15).
For any particular configuration of the antennas of transmitter 24. T ₀ may be determined by either of two different calibration procedures.
In the experimental calibration procedure, probe 10 is oriented so that T is a unit matrix, probe 10 is moved to a succession of positions relative to transmitter 24, and M is measured at each position. The equation $T_{0} = {Pd}^{- 1} P^{T} M$

gives To at each of those calibration positions.
There are two variants of the theoretical calibration procedure, both of which exploit reciprocity to treat receiver 14 as a transmitter and transmitter 24 as a receiver. The first variant exploits the principle of reciprocity. The sensor elements are modeled as point sources, including as many terms in their multipole expansions as are necessary for accuracy, and their transmitted magnetic fields in the plane of transmitter 24 are calculated at a succession of positions relative thereto, also with probe 10 oriented so that T is a unit matrix. The EMF induced in the antennas of transmitter 24 by these time-varying magnetic fields is calculated using Faraday's law. The transfer function of reception circuitry 34 then is used to compute M at each calibration position, and equation (16) gives T ₀ at each calibration position. In the second variant, the magnetic field generated by each antenna of transmitter 24 at the three frequencies ω₁, ω₂ and ω₃ is modeled using the Biot-Savart law. Note that each frequency corresponds to a different sensor 16, 18 or 20. The signal received at each sensor is proportional to the projection of the magnetic field on the sensitivity direction of the sensor when object 10 is oriented so that T is a unit matrix. This gives the corresponding column of M up to a multiplicative constant and up to a correction based on the transfer function of reception circuitry 34.
To interpolate T ₀ at other positions, a functional expression for T ₀ is fitted to the measured values of T ₀. Preferably, this functional expression is a polynomial. It has been found most preferable to express the Euler angles α,β and y that define T ₀ as the following 36-term polynomials. The arguments of these polynomials are not direct functions of Cartesian coordinates x, y and z, but are combinations of certain elements of matrix W that resemble x, y and z, specifically, a = W ₁₃/(W ₁₁+W ₃₃), which resembles x; b = W₂₃ /W₂₂+W₃₃), which resembles y, and c = log(1/W ₃₃), which resembles z. Using a direct product notation, the 36-term polynomials can be expressed as: $α = (a a^{3} a^{5}) (b, b^{3}, b^{5}) (1, c, c^{2}, c^{3}) AZcoe$
$β = (a a^{3} a^{5}) (1, b^{2}, b^{4}, b^{6}) (1, c, c^{2}) ELcoe$
$γ = (1 a^{2} a^{4} a^{6}) (b, b^{3}, b^{5}) (1, c, c^{2}) RLcoe$

where AZcoe, ELcoe and RLcoe are 36-component vectors of the azimuth coefficients, elevation coefficients and roll coefficients that are fitted to the measured or calculated values of the Euler angles. Note that to fit these 36-component vectors, the calibration procedure must be carried out at at least 36 calibration positions. At each calibration position, W is computed from M using equation (13), and the position-like variables a, b and c are computed from W as above.
Similarly, the Cartesian coordinates x, y and z of probe 10 relative to the reference frame of transmitter 24 may be expressed as polynomials. It has been found most preferable to express x, y and z as the, following 36-tern polynomials: $x = (a a^{3} a^{5}) (1, b, b^{4}) (1, c, c^{2}, c^{3}) Xcoe$
$y = (1 a^{2} a^{4}) (b, b^{3}, b^{5}) (1, c, c^{2}, c^{3}) Ycoe$
$z = (1 a^{2} a^{4}) (1, b^{2}, b^{4}) (1, d, d^{2}, d^{3}) Zcoe$

where Xcoe, Ycoe and Zcoe are 36-component vectors of the x-coefficients, the y-coefficients, and the z-coefficients, respectively; and d = log(c). As in the case of the Euler angles, these position coordinate coefficients are determined by either measuring or computing M at at least 36 calibration positions and fitting the resulting values of a, b and c to the known calibration values of x, y and z. Equations (17) through (22) may be used subsequently to infer the Cartesian coordinates and Euler angles of moving and rotating probe 10 noniteratively from measured values of M.
Figures 7A, 7B and 7C show three alternative configurations of paired adjacent loop antennas 26' and 28'. The arrows indicate the direction of current flow that emulates a single loop antenna coincident with the outer perimeter of antennas 26' and 28'. Other useful coplanar overlapping antenna configurations are described in PCT Publication No. WO 96/03188 , entitled "Computerized game Board", which is incorporated by reference for all purposes as if fully set forth herein.
Figure 8 is a schematic block diagram of driving circuitry 32 for driving a generic antenna 25 that represents any one of loop antennas 26, 28 or 30. A digital signal generator 50 generates samples of a sinusoid that are converted to an analog signal by a D/A converter 52. This analog signal is amplified by an amplifier 54 and sent to the positive input 60 of a differential amplifier 58. Loop antenna 25 is connected both to the output 64 of differential amplifier 58 and to the negative input 62 of differential amplifier 58. Negative input 62 also is grounded via a resistor 66. The feedback loop thus set up drives antenna 25 at the frequency of the sinusoid generated by signal generator 50, and makes antenna 25 appear to be an open circuit at all other frequencies.
Unlike the circuitry of WO 97/36143 , which acts to offset the influence of one loop antenna on another, the circuitry of Figure 8 decouples loop antenna 25 from the other loop antennas. The advantage of the present invention over WO 97/36143 is evident. Consider, for example how WO 97/36143 and the present invention correct for the mutual inductances of loop antenna 26, radiating at a frequency ω₁, and loop antenna 30; radiating at a frequency ω_2. The goal is to set up the field of frequency ω₁, that would be present if only loop antenna 26, and not loop antenna 30, were present, and to set up the field of frequency ω₂ that would be present if only loop antenna 30, and not loop antenna 26, were present. By Faraday's and Ohm's laws, the time rate of change of the magnetic flux through loop antenna 26 is proportional to the current through loop antenna 26, and the time rate of change of the magnetic flux through loop antenna 30 is proportional to the current through loop antenna 30. In the absence of loop antenna 30, loop antenna 26 sets up a certain time-varying magnetic flux of frequency ω₁ across the area that would be bounded by loop antenna 30 if loop antenna 30 were present. The method of WO 97/36143 forces the time rate of change of this magnetic flux through loop antenna 30 to be zero. Because the magnetic flux has no DC component, the magnetic flux itself through loop antenna 30 therefore also vanishes, which is contrary to the situation in the absence of loop antenna 30. By contrast, the present application makes loop antenna 30 appear to be an open circuit at frequency ω₁ and so does not change the magnetic flux from what it would be in the absence of loop antenna 30.
Figure 9 shows, schematically, a C-mount fluoroscope 80 modified for simultaneous real-time image acquisition and intrabody navigation. Fluoroscope 80 includes the conventional components of a C-mount fluoroscope: an x-ray source 82 and an image acquisition module 84 mounted on opposite ends of a C-mount 78, and a table 86 whereon the patient lies. Image acquisition module 84 converting x-rays that transit the patient on table 86 into electronic signals representative of a 2D image of the patient. C-mount 78 is pivotable about an axis 76 to allow the imaging of the patient from several angles, thereby allowing the reconstruction of a 3D image of the patient from successive 2D images. In addition, either a receiver 114, similar to receiver 14, or transmitter 24, is rigidly mounted on C-mount 78. Receiver 114 or transmitter 24 serves to define a frame of reference that is fixed relative to C-mount 78. The other components shown in Figure 1, i.e., driving circuitry 32, reception circuitry 34, and control/processing unit 36, are connected to transmitter 24 and to receiver 14 in probe 10 as described above in connection with Figure 1. In addition, signals from receiver 114 that correspond to the electromagnetic waves generated by transmitter 24' are sent to reception circuitry 134 that is identical to reception circuitry 34, and controller/processor 36 directs the reception of received signals by reception circuitry 134 and the acquisition of an image of the patient by image acquisition module 84 of fluoroscope 80.
By determining the position and orientation of probe 10 relative to the frame of reference defined by transmitter 24, controller/processor 36 determines the position and orientation of probe 10 relative to each acquired 2D image. Alternatively, the electromagnetic signals are transmitted by a transmitter 24' that is not attached to C-mount 78, and controller/processor 36 determines the position and orientation of probe 10 relative to the 2D images by determining the positions and orientations of receivers 14 and 114 relative to transmitter 24'. Controller/processor 36 synthesizes a combined image that includes both the 3D image of the patient acquired by fluoroscope 80 and an icon representing probe 10 positioned and oriented with respect to the 3D image of the patient in the same way as probe 10 is positioned and oriented with respect to the interior of the patient. Controller/processor 36 then displays this combined image on a monitor 92.
C-mount fluoroscope 80 is illustrative rather than limitative. The application includes all suitable devices for acquiring 2D or 3D images of the interior of a patient, in modalities including CT, MRI and ultrasound in addition to fluoroscopy.
Under certain circumstances, the image acquisition and the intrabody navigation may be done sequentially, rather than simultaneously. This is advantageous if the medical imaging facilities and the medical treatment facilities can not be kept in the same location. For example, the human skull is sufficiently rigid that if a receiver of the present invention is rigidly mounted on the head of a patient using an appropriate headband, then the position and orientation of the receiver is a sufficient accurate representation of the position and orientation of the patient's head to allow intracranial navigation. Figure 11 shows a head 94 of a patient inside a (cut-away) CT scanner 98. As in the case of fluoroscope 80 of Figure 9, receiver 114 and transmitter 24 are rigidly attached to CT scanner 98, transmitter 24 being so attached via an arm 100. CT scanner 98 acquires 2D x-ray images of successive horizontal slices of head 94. A receiver 214 is rigidly mounted on head 94 using a headband 96. As the 2D images are acquired, the position and orientation of receiver 214 with respect to each image is determined by the methods described above for determining the position and orientation of probe 10 with respect to the 2D images acquired by fluoroscope 80. These positions and orientations are stored, along with the 2D images, in control/processing unit 36. Subsequently, during medical treatment of head 94 that requires navigation of probe 10 through head 94, the position and orientation of probe 10 in head 94 is determined using signals from receivers 14 and 214 in the manner described above for positioning and orienting probe 10 with respect to C-mount 78 of fluoroscope 80 using receivers 14 and 114. Given, now, for each 2D CT image, the position and orientation of probe 10 with respect to receiver 214 and the position and orientation of receiver 214 with respect to that 2D image, it is trivial to determine the position and orientation of probe 10 with respect to that 2D image. As in the case of the simultaneous imaging and navigation depicted in Figure 9, controller/processor 36 now synthesizes a combined image that includes both the 3D image of head 94 acquired by CT scanner 98 and an icon representing probe 10 positioned and oriented with respect to the 3D image of head 94 in the same way as probe 10 is positioned and oriented with respect to head 94. Controller/processor 36 then displays this combined image on monitor 92.
As in the case of fluoroscope 80, CT scanner 98 is illustrative rather than limitative. The application includes all suitable devices for acquiring 2D or 3D images of a limb of a patient, in modalities including MRI, ultrasound and fluoroscopy in addition to CT. Note that this method of image acquisition followed by intrabody navigation allows the a centrally located imaging device to serve several medical treatment facilities.
Figure 14 is a partially exploded, partial perspective view of a modified C-mount fluoroscope 80'. Like C-mount fluoroscope 80, C-mount fluoroscope 80' includes an x-ray source 84 and an image acquisition module 82 at opposite ends of a C-mount 78. Image acquisition module 82 includes an image intensifier 83, a front face 85 whereof faces x-ray source 84, and a CCD camera 87, mounted on the end of image intensifier 83 that is opposite front face 85, for acquiring images that are intensified by image intensifier 83. Image intensifier 83 is housed in a cylindrical housing 91. In addition, fluoroscope 80' includes an annular compensator 500 made of a magnetically permeable material such as mu-metal.
The need for compensator 500 derives from the fact that front face 85 is electrically conductive. The electromagnetic waves generated by transmitter 24 or 24' induce eddy currents in front face 85 that distort the electromagnetic field sensed by receiver 14. Placing a mass of a magnetically permeable substance such as mu-metal in the proper spatial relationship with front face 85 suppresses this distortion. This is taught, for example, in U. S. Patent 5,760,335, to Gilboa , which patent is incorporated by reference for all purposes as if fully set forth herein, in the context of shielding a CRT from external radiation without perturbing the electromagnetic field external to the CRT.
Preferably, compensator 500 is a ring, 5 cm in axial length, of mu metal foil 0.5 mm thick. Compensator 500 is slidably mounted on the external surface 89 of cylindrical housing 91, as indicated by double-headed arrows 504, and is held in place by friction. It is straightforward for one ordinarily skilled in the art to select a position of compensator 500 on housing 91 that provides the optimal suppression of distortions of the electromagnetic field outside image intensifier 83 due to eddy currents in front face 85.
It often is desirable to retrofit a new apparatus such as receiver 14 to an existing catheter rather than to design a new probe 10 that includes both the new apparatus and the functionality of an already existing probe. This retrofit capability is particularly important if probe 10 would have been used for medical applications, and both the apparatus and the existing probe had already been approved for medical applications by the relevant regulatory bodies. Such a retrofit capability then would preclude the need to obtain regulatory approval for the new probe, a process that often is both expensive and time-consuming.
Figure 16 illustrates just such a retrofit capability, for adapting a satellite 550 to a substantially cylindrical catheter 552 for invasively probing or treating a body cavity such as a chamber of the heart. Satellite 550 is an instrumentation capsule that may contain receiver 14 or any other medically useful apparatus. For example, satellite 550 may contain an apparatus for ablating cardiac tissue. A catheter such as catheter 552 is introduced to the body cavity of a patient via the patient's blood vessels, via an introducer sheath. It is important that the external diameter of the introducer sheath be minimized, to reduce the risk of bleeding by the patient. Consequently, the external diameter of catheter 552 also must be minimized, and any scheme for retrofitting satellite 550 to catheter 552 must allow satellite 550 to be introduced into the introducer sheath along with catheter 552. It is the latter requirement that generally precludes simply attaching satellite 550 to catheter 552. In addition, if satellite 550 includes receiver 14, with the intention of using receiver 14 to track the position and orientation of catheter 550, then, when satellite 550 and catheter 552 are deployed within the body cavity, satellite 550 must have a fixed position and orientation relative to catheter 552.
The retrofitting scheme of Figure 16 achieves these ends by providing satellite 550 and catheter 552 with a mechanism for providing only a loose mechanical connection between satellite 550 and catheter 552 as satellite 550 and catheter 552 are introduced to the body cavity, and only then securing satellite 550 to catheter 552 at a fixed position and orientation relative to catheter 552. Figure 16A shows a thin flexible tether 554 attached to proximal end 556 of satellite 550. Tether 554 provides a mechanical link to the outside of the patient. Depending on the instrumentation installed in tether 554, tether 554 may also provide a communications link to the outside of the patient. For example, if satellite 550 includes receiver 14, then extensions of wire pairs 38 are included in tether 554. Rigidly attached to tether 554 is a hollow cylindrical sleeve 558 whose inner diameter is the same as the outer diameter of catheter 552.
The remainder of the mechanism for reversibly securing satellite 550 to catheter 552 is shown in Figure 16B. Catheter 552 is provided, near distal end 564 thereof, with a pocket 560 made of a flexible, resilient, elastic material. Pocket 560 is attached rigidly to the outer surface of catheter 552. Pocket 560 includes an aperture 562, which is adjacent catheter 552 at the proximal end of catheter 552, and which accommodates tether 554. Pocket 560 is sized to accommodate satellite 550 snugly therein via an opening in distal end 566 of pocket 560.
Satellite 550, catheter 552 and the associated securing mechanism are assembled as shown in Figure 16C, with tether 554 running through aperture 562, sleeve 558 encircling catheter 552 proximal of pocket 560. and satellite 550 distal of pocket 560. Catheter 552 and tether 554 are shown emerging from the distal end of a protective jacket 568. Preferably, sleeve 558 is made of a low-friction material such as Teflon^™, to allow sleeve 558 to slide freely along catheter 552. The assembly shown in Figure 16C is introduced to the introducer sheath with satellite 550 in front of catheter 552. During this introduction, pocket 560 is compressed against the outer surface of catheter 552 by the introducer sheath. Tether 554 is sufficiently flexible to bend along with catheter 552 and jacket 568 as the assembly shown in Figure 16C passes through the patient's blood vessels, but is sufficiently rigid to push satellite 550 ahead of distal end 564 of catheter 552 as catheter 552 is inserted into the patient. As a result, satellite 550 and distal end 564 of catheter 552 reach interior of the targeted body cavity in the configuration illustrated in Figure 16C. At this point, pocket 560 opens, and tether 554 is pulled to withdraw satellite 550 into pocket 560 via the opening in distal end 566 of pocket 560. Satellite 550 and tether 554 now are held by pocket 560, sleeve 558 and jacket 568 in a fixed position and orientation relative to catheter 552, as illustrated in Figure 16D.
Subsequent to treatment, tether 554 is pushed to restore the configuration shown in figure 16C, to allow catheter 552 and satellite 550 to be withdrawn from the patient.
While the invention has been described with respect to a limited number of embodiments, it will be appreciated that many variations, modifications and other applications of the invention may be made within the scope defined by the appended claims.

Claims

A system for tracking a position and an orientation of a probe (10), comprising:
three sensors (16, 18, 20') each for detecting a different component of a vector force field, two sensors (16, 18) of said three sensors including two sensor elements (16a, 16b, 18a, 18b) disposed symmetrically about a common reference point (22) in the probe (10), said three sensors (16, 18, 20) being mounted inside the probe (10), wherein

another sensor (20') of said three sensors is centered on said common reference point (22), and said sensor elements (16a, 16b, 18a, 18b) and said other sensor (20') are disposed collinearly along a longitudinal axis (12) of the probe (10) characterized in that

each of said sensor elements (16a, 16b, 18a, 18b) exclusively consists of a coil of conducting wire wound on a respective ferrite core (70), and said other sensor (20') exclusively of a single coil of conducting wire wound on a ferrite core, and
wherein said single coil of said other sensor (20') is oriented parallel to said longitudinal axis (12), and said coils of said sensor elements are oriented perpendicular to said longitudinal axis (12).
The system of claim 1, wherein the coils of each sensor element pair are mutually parallel, coils of one sensor element pair being perpendicular to coils of the other sensor element pair.